Research And Implementation Of Text Classification Algorithm For NQI Comprehensive Service

National quality infrastructure provides important support for the healthy stable development of the economy and society as a framework which is required to establish standards for measurement,inspection and accreditation in country's need.This thesis relies on the NQI key special sub-project "NQI comprehensive service information strategy and information system research and development".With respect to the problem that the current domestic NQI comprehensive service capabili ty is weak and the open extent is relatively low,the thesis mainly focuses on the work about the automatic collection of NQI resource information,classification methods and the design of the NQI resource information classification system to provide data support for the NQI integrated service information system.The main research contents of this thesis are as follows:(1)Investigate the basic situation and service methods of domestic measurement,standard,inspection and certification service resources,point out the shortcomings in the NQI service process,analyze the needs of NQI text classification and the characteristics of NQI text,and propose a technical solution for NQI text classification.(2)Collect the measurement,inspection and certification service information of specific websites through web crawler,formulate NQI text storage specifications and determine the fields required for classification.The collected text is cleaned and tagged to construct the NQI text data set.Afterwards the text is converted into word vector matrix form as input to deep learning model.(3)Propose an NQI resource text classification algorithm based on the attention mechanism of long and short-term memory network,and improve the preprocessing and key feature enhancement of NQI text.Collect the inspection and testing service text information of 10 industries including food,chemical,automobile,clothing and leather,environment,electronics and electrical appliances as the text data set to be classified,and conduct text classification experiments and related comparative experiments.Results show that the classification algorithm proposed in this thesis has better performance,with a classification accuracy of 98.7%(4)Design the NQI text classification system wit h the algorithm proposed in this thesis.The system is equipped with functions of NQI information collection,storage,text classification and display of NQI research results.All functions of the system are tested.The test results show that the functions of the NQI resource information classification system designed in this thesis meet the data requirements of the NQI integrated service information system.